Background. Although the use of antibiotics in the treatment of acute exacerbation of chronic obstructive pulmonary disease (COPD) is largely accepted, controversy remains regarding whether the choice of antibiotic has any impact on outcome. Our aim was to compare the effects of the combination of trimethoprim and sulfamethoxazole and ciprofloxacin in patients treated for severe COPD exacerbation requiring mechanical ventilation.
Methods. In a randomized, double-blind trial, we included 170 patients with an acute exacerbation of COPD requiring mechanical ventilation. Enrolled patients received trimethoprim-sulfamethoxazole (n = 85) or ciprofloxacin (n = 85) for 10 days. Main outcomes were hospital death and need for an additional course of antibiotics. Secondary outcomes were duration of mechanical ventilation, length of hospital stay, and exacerbation-free interval.
Results. Combined hospital death and additional antibiotic prescription rates were similar in the 2 groups (16.4% vs 15.3% for trimethoprim-sulfamethoxazole group vs ciprofloxacin group; difference, 1.1%; 95% confidence interval [CI] −9.8% to 12.0%; P = .832). Hospital death occurred in 7 patients (8.2%) receiving trimethoprim-sulfamethoxazole and 8 patients (9.4%) receiving ciprofloxacin (difference, −1.2%; 95% CI, −9.7 to 7.3; P = .90). The need for an additional antibiotic course was observed in 8 patients in the trimethoprim-sulfamethoxazole group and 5 patients in the ciprofloxacin group (difference, 2.3%; 95% CI, −5.4 to 10.0; P = .549). The mean exacerbation-free interval (± standard deviation) was similar in both treatment groups (83 ± 25 vs 79 ± 22 for the trimethoprim-sulfamethoxazole group vs ciprofloxacin group; difference, 4 days; 95% CI, −15 to 19 days; P = .41). Duration of mechanical ventilation and hospital stay was not significantly different between the 2 groups.
Conclusions. In patients with acute exacerbation of COPD requiring mechanical ventilation, efficacy of trimethoprim-sulfamethoxazole was not inferior to ciprofloxacin. Trial registration. ClinicalTrials.gov identifier: NCT00791505
Antibiotic treatment of chronic obstructive pulmonary disease (COPD) exacerbation has become widely accepted in severe cases [1–3]. Use of amoxicillin, the combination of trimethoprim and sulfamethoxazole, or tetracyclines has long been regarded as standard therapy for patients with acute exacerbation of COPD [4]. Although emerging resistance of common respiratory pathogens to these old classes of antibiotics might limit their recommendation in acute exacerbation of COPD, available data failed to demonstrate clear superiority of newer antimicrobial agents. Moreover, most clinical trials comparing new with standard antibiotics were retrospective or conducted in outpatients who, in most cases, would have never required antibiotics [5–7]. Accordingly, selective inclusion of patients with severe acute exacerbation of COPD who have a high incidence of bacterial infection is most suitable for such a comparison [3, 8, 9]. Regarding its excellent activity against gram-negative pathogens, especially Pseudomonas strains, ciprofloxacin constitutes a suitable choice in treating severe acute exacerbation of COPD [10, 11]. Moreover, ciprofloxacin offers substantial economic benefits compared with firstline agents in the treatment of acute exacerbation of COPD [12]. Therefore, we designed a controlled, double-blind study in intensive care unit (ICU) patients to compare trimethoprim-sulfamethoxazole with ciprofloxacin in the treatment of patients admitted to the ICU with acute exacerbation of COPD requiring mechanical ventilation.
Design overview. This was a prospective, randomized, double-blind study involving patients with acute exacerbation of COPD requiring mechanical ventilation conducted in compliance with the institutional review board regulations. Informed consent was obtained from each patient. The first patient entered the study in July 2002, and the last patient completed the study in June 2005.
Setting and participants. Adult patients (age, ⩾40 years) admitted to 2 medical ICUs (Fattouma Bourguiba University Hospital in Monastir, Tunisia, and Tahar Sfar University Hospital in Mahdia, Tunisia) because of acute exacerbation of COPD were enrolled in the study. The diagnosis of acute exacerbation of COPD required a history of COPD with clinical evidence of a purulent bronchitis in addition to acute respiratory failure requiring mechanical ventilation within the first 24 h after ICU admission. Acute respiratory failure was defined as an association of dyspnea with at least 2 of the following characteristics: respiratory rate, >30 breaths/min; arterial partial pressure of carbon dioxide, >6 kPa; and arterial pH, <7.30 just before the initiation of mechanical ventilation. A chest radiograph confirming the absence of pulmonary findings consistent with pneumonia was required.
Patients with the following conditions were excluded: hypersensitivity to quinolones or trimethoprim-sulfamethoxazole or presence of a significant comorbid condition that precluded evaluation of the therapeutic response (eg, hepatic or renal disease, human immunodeficiency virus infection, leukopenia with a blood cell count <3000 cells/mm3, current use of immunosuppressive or chemotherapeutic agents, or malignant tumor). Patients with the following conditions were also excluded: gastrointestinal tract abnormalities that would have prevented absorption of oral medication, patients with concomitant infection requiring systemic antibacterial treatment, and patients who received systemic antimicrobial treatment within 10 days before study drug administration.
Randomization and interventions. Baseline evaluation included verification of inclusion and exclusion criteria, informed consent, medical history, physical examination (including vital signs and the simplified acute physiology score calculated in the first 24 h after ICU admission) [13], electrocardiography, chest radiography, and microbiology testing. Included patients were systematically assigned to noninvasive ventilation. Intubation was indicated in patients unlikely to benefit from noninvasive ventilation, including those with hypotension, uncontrolled cardiac ischemia, inability to protect the airway, excessive secretion, agitation or uncooperativeness, or a condition that precludes the placement of the mask.
Included patients were randomized to receive either trimethoprim- sulfamethoxazole (160/800 mg twice daily; Galpharma Laboratories) or ciprofloxacin (750 mg twice daily; Teriak Laboratories) for 10 days. A double-dummy system was used such that matching comparative placebos were given concurrently with the active drug during the treatment period. The randomization scheme was generated by a random-number generator with a block size of 5. Drug kits with identical appearance were prepared in the pharmacy department where they were stored. Access to randomization codes and drug kit identity codes was restricted to the head pharmacist of the hospital. Patients and ICU staff were masked to treatment code until completion of the study.
All patients received enteral nutrition orally or by nasogastric or orogastric tube. After randomization an attempt was made to obtain a tracheal aspiration sample or a sputum sample from each patient. These samples were Gram stained, and samples that revealed a preponderance of neutrophils were cultured for bacterial pathogens. In intubated patients, infection was defined only when quantitative cultures yielded ⩾105 CFU/mL of bacterial pathogens. All bacteria considered to be pathogenic were identified and tested for susceptibility to trimethoprim-sulfamethoxazole and ciprofloxacin according to standard techniques. Long-term follow-up was made by scheduled visits at months 1, 3, and 6 after study treatment. Patients who missed their visit were contacted directly or through their relatives and general practitioners when needed. The time to next acute exacerbation and the exacerbation-free interval (EFI) were recorded.
Outcomes and measurements. Outcome analyses were performed on the intent-to-treat population. The primary outcome analysis was combined events (hospital death and need for an additional course of antibiotics). Secondary outcomes were EFI, duration of mechanical ventilation, and length of hospital stay. The decision to initiate new antibiotics was left to the discretion of the treating physician, and if resistant species were cultured, the protocol treatment was not systematically changed unless the clinical course of the patient worsened. The cause of any new course of antibiotic was reported and clearly notified. Study treatment was prematurely discontinued if major adverse effects occurred or if the patient's condition deteriorated because of worsening or persistence of clinical symptoms of infection requiring additional antibiotics. Adverse effects were reported and rated by the investigator as possibly, probably, or definitely related to the study treatment.
Statistical analysis. The intent-to-treat population included all randomized patients receiving at least 1 dose of study drug. The study was powered as a noninferiority study, and the sample size was calculated to demonstrate that trimethoprim-sulfamethoxazole was not >10% less effective than ciprofloxacin. Sample size was based on a predicted failure rate (combining hospital death and need for additional antibiotic rate) of 15% in the ciprofloxacin arm with <10% variation between study arms (α = .05 [2 sided] and α = .20). On this basis, 85 patients were needed per treated group.
Differences between treatment groups in the distribution of demographic variables and baseline characteristics were tested using 2-way analysis of variance for continuous variables with the Cochran-Mantel-Haenszel test for categorical variables. The Cochran-Mantel-Haenszel test was also used to analyze selected adverse events.
The 2-sided 95% confidence interval (CI) for a difference in treatment response rates was used to determine equivalence between trimethoprim-sulfamethoxazole and ciprofloxacin for efficacy variables. Time to occurrence of first relapse (EFI) was analyzed in each treatment group by Kaplan-Meier survival curves and compared between treatment groups by the log rank test. All statistical analyses were performed using SPSS statistical software, version 11.0 (SPSS).
A total of 357 patients were enrolled in the study, 170 of whom were randomized. The randomized and nonrandomized populations were comparable in age, sex, mean duration of chronic bronchitis, and number of acute exacerbation of COPD episodes in the previous year. A flowchart showing randomization of patients into the study is shown in Figure 1. The most frequent reasons for nonrandomization were prior use of antibiotics (n = 58) and pneumonia (n = 49). Description of baseline characteristics and clinical presentation of the patients are given in Table 1. No statistically significant differences were found between the trimethoprim-sulfamethoxazole (n = 85) and ciprofloxacin (n = 85) groups with respect to demographic characteristics, previous pulmonary function, number of exacerbations requiring hospitalization during the previous year, laboratory tests, and signs of infection at baseline. The 2 groups were also similar with respect to initial severity of the episode of acute exacerbation as assessed by the simplified acute physiology score calculated in the first 24 h of ICU admission and blood gas values.
Kaplan-Meier survival analysis of patients without relapse of chronic obstructive pulmonary disease exacerbation in the intention-to treat patients in the ciprofloxacin (hatched line) and trimethoprim-sulfamethoxazole (solid line) groups. No significant difference was observed.
Bacteria were grown from sputum, tracheal aspiration, or both in 73 patients (43%). A similar proportion of microorganisms could be cultured in both groups: 38 microorganisms (45%) and 35 microorganisms (41%) were identified in the trimethoprim-sulfamethoxazole and ciprofloxacin groups, respectively. Table 2 gives detailed findings of bacterial growth from sputum and tracheal aspiration in the 2 groups. The most frequently isolated bacteria were Streptococcus pneumoniae (n = 18), Haemophilus influenzae (n = 17), and Pseudomonas aeruginosa (n = 13). Four strains of H. influenzae were resistant to trimethoprim-sulfamethoxazole, but all isolates were susceptible to ciprofloxacin. The rate of S. pneumoniae resistance was not significantly different between study groups (3 [38%] and 3 [30%] in the trimethoprim-sulfamethoxazole and ciprofloxacin groups, respectively).
Primary outcome. Combined hospital death and additional antibiotic prescription rates were similar in the 2 groups (16.4% vs 15.3% in the trimethoprim-sulfamethoxazole vs. ciprofloxacin group; difference, 1.1%; 95% CI, −9.8% to 12.0%; P = .832) (Table 3). During the study 15 patients died in the hospital, 7 (8.2%) in the trimethoprim-sulfamethoxazole group and 8 (9.4%) in the ciprofloxacin group (difference, −1.2%; 95% CI, −9.7% to 7.3%; P>.05). Causes of death were septic shock (n = 9), adult respiratory distress syndrome (n = 2), acute myocardial infarction (n = 1), pulmonary embolus (n = 1), and unexpected cardiac arrest (n = 2). All septic shock deaths were related to secondary nosocomial pneumonia; in only 1 patient initial bacterial assessment found a resistant strain (S. pneumoniae in the trimethoprim-sulfamethoxazole group). The mean ±SD time delay from ICU admission to death was 8.7±4.1 days. The rate of additional antibiotic prescriptions was similar in both treatment groups (8.2% vs 5.9% in the trimethoprim-sulfamethoxazole vs ciprofloxacin group; difference, 2.3%; 95% CI, −5.4% to 10.0%; P = .549). Antibiotic prescriptions included imipenem (n = 4), ceftazidime (n = 3), and cefotaxime (n = 2). No correlation was found between antibiotic resistance and treatment failure. There were 7 drug-resistant strains (70%) in nonresponders and 13 (26%) in responders (n = .78). In the nonresponder group (patients died or required additional antibiotic), bacterial resistance was observed in 3 isolates in the trimethoprim-sulfamethoxazole group (H. influenzae, 1; other Streptococcus species, 1; Enterobacter species, 1) and in 2 isolates in the ciprofloxacin group (S. pneumoniae, 1; P. aeruginosa, 1).
Secondary outcomes. The mean EFI was similar in both treatment groups (83±25 days vs 79±22 days in the trimethoprim-sulfamethoxazole vs. ciprofloxacin group; difference, 4 days; 95% CI, −15% to 19%; P = .41). Kaplan-Meier curves for EFI were similar for the 2 treatment groups (Figure 2).
Of 38 patients initially receiving noninvasive ventilation in the trimethoprim-sulfamethoxazole group, 17 (45%) were secondarily intubated versus 13 (34%) in the ciprofloxacin group (P = .347). Duration of mechanical ventilation and length of hospital stay were similar in the 2 study groups.
Adverse events were minor and comparably distributed in both treatment groups (Table 3). Most of these adverse events were considered by the investigators to be possibly related to the study medication. One patient in the trimethoprim-sulfamethoxazole group experienced maculopapular rash that led to discontinuation from the study by the investigator; this patient was treated with a combination of amoxicillin and clavulanate and discharged alive thereafter. No renal, hepatic, or hematologic reaction developed in either group of patients. All patients received their full dose of assigned protocols antibiotics.
Our findings indicate that administration of trimethoprim-sulfamethoxazole is noninferior to ciprofloxacin in patients with acute exacerbation of COPD requiring mechanical ventilation. This noninferiority was demonstrated for both the primary outcome of combined events (hospital death and need for additional antibiotic treatment) and the secondary outcomes of mechanical ventilation duration, length of hospital stay, and EFI. Adverse events were reported in a similar proportion of patients in both groups. No serious adverse effects were observed in the 2 groups of patients, and tolerance was generally good.
Use of simple antibiotics, such as amoxicillin, trimethoprimsulfamethoxazole, or tetracycline, has long been considered standard therapy for acute exacerbation of COPD [1]. However, increasing emergence of resistance to these traditional antibiotics in common respiratory pathogens has become a real challenge in the choice of the suitable and adequate antibacterial agents [14]. However, most available prospective trials failed to demonstrate superiority of newer agents over older classes of respiratory antibiotics. This is likely because these activecontrol studies did not include patients with enough severity because this would enhance the likelihood of a bacterial origin of acute exacerbation of COPD. Our findings are in contrast with those of Wilson et al [6], who concluded that moxifloxacin, a third-generation quinolone active against S. pneumoniae, is superior to standard therapy. However, this study was criticized because the authors pooled the 3 comparative antibiotics (cefuroxime-axetil, amoxicillin, and clarithromycin) in the same group. Similar concerns could be addressed about a recent meta-analysis that suggested that first-line antimicrobial agents were associated with less treatment success compared with second-line and newer agents [7]. Moreover, the interesting outcome of COPD EFI was not evaluated in most of the studies included in this meta-analysis. Undoubtedly, EFI is of paramount clinical importance because frequent exacerbations are associated with a faster decline in pulmonary function [15], increased costs [16], and impairment quality of life [17]. Our findings support those of Peng et al [18], who compared the EFI in 2 antibiotic groups, categorized as the first-line agent group (amoxicillin, cephalexin, doxycycline, erythromycin, and trimethoprim-sulfamethoxazole) and second-line agent group (amoxicillin-clavulanate, azithromycin, cefuroxime, clarithromycin, levofloxacin, and ofloxacin). They found no significant difference between both groups with regard to EFI (33.4±19.7 weeks vs 32.9±19.6 weeks in the first-line vs second-line antibiotics group; P = .73). In another study, the percentage of COPD patients who remained free from relapse for 6 months was 34% for an older antibiotic versus 28% for newer antibiotic regimen (P = .37) [19].
There are several limitations to our study. First, several investigators prefer new quinolones over ciprofloxacin in the treatment of acute exacerbation of COPD. Unfortunately, available data from randomized controlled trials are not enough to support this tendency. In addition, like most new quinolones, ciprofloxacin performs well against H. influenza and Moraxella catarrhalis because its therapeutic target minimum inhibitory concentration is well above the minimum inhibitory concentration required to inhibit the growth of 90% of these organisms. Against S. pneumoniae, ciprofloxacin activity is lower than that of new-generation quinolones but it is reasonably acceptable [20]. Moreover, in patients with acute exacerbation of COPD with a history of severe chronic bronchitis, ciprofloxacin treatment seems to offer substantial clinical and economic benefits [12]. Second, the absence of difference between both treatments should raise the possibility that both antibacterial agents are ineffective. This is understandable in outpatients with low major event rates but probably not in severe patients requiring mechanical ventilation because the clinical outcome would be worse than that seen in the present study. Indeed, we previously demonstrated that in this category of patients ofloxacin has a significant beneficial effect compared with placebo [3]. Moreover, the consistency of short and late outcomes in our study increases the validity of our results. Third, a bacterial origin of acute exacerbation of COPD was not confirmed in all of our patients, but the number of cultures that were positive for respiratory pathogens in our patients was similar to that anticipated in previous studies [21]. In addition, the high levels of serum C-reactive protein recorded in almost all of our patients might support the predominance of bacterial origin of their exacerbation [22]. Fourth, analysis of subsequent bacterial isolates and eradication rates with resistance patterns cannot be assessed in our patients based on the current study. Nonetheless, the use of objective clinical outcomes, such as death and EFI, could circumvent in many respects the controversial interpretation of microbiological findings in patients with COPD [23, 24].
In conclusion, the results of this study show that trimethoprim-sulfamethoxazole and ciprofloxacin are clinically equally effective and exhibit similar safety profiles in patients with severe exacerbation of COPD requiring mechanical ventilation. Our findings highlight the necessity to reappraise the value of standard antibiotic therapy, such as trimethoprim-sulfamethoxazole, with regard to their low cost and their ability to decrease the use of new and broad-spectrum agents. Alternating the use of standard antibiotics with several different classes of new antimicrobial agents on a regular basis could be an effective way to limit bacterial resistance in ICU and non-ICU settings.
Financial support. This research was conducted as part of a Research Unit 04/UR/08-20 activity funded by the Tunisian Ministry of Higher Education, Technology and Scientific Research.
Potential conflicts of interest. All authors: no conflicts.
IDSA Members: For your free access to this journal, log in via the IDSA members area.
Open access options for authors visit Oxford Open
This journal enables compliance with the NIH Public Access Policy